80% of Enterprises Run AI Agents in Production, Yet Security Confidence Is 2.32/5 — Team8 Puts $365M to Work

For SaaS and cloud infrastructure leaders, the numbers are alarming: 80% of organizations already have AI agents in production, but on a scale of 5, their average confidence in securing them is a dismal 2.32. Team8’s new fund is a direct answer to this existential risk, promising to shepherd a new class of AI-native software infrastructure and security startups that can embed governance, visibility, and threat protection into the fabric of AI-augmented applications. This isn’t a niche; it’s becoming mandatory middleware for every enterprise SaaS deployment.

Team8, the global venture fund, has announced $365 million in new capital, comprising a $265 million third fund and over $100 million earmarked for follow-on investments in its most promising portfolio companies. This latest raise pushes Team8’s total assets under management to nearly $2 billion across eight funds since its founding in 2014. The announcement, made on August 11, 2026, underscores the firm’s conviction that the next wave of enterprise value will be created by AI-native startups that can bridge the gap between accelerating AI innovation and the practical security, governance, and infrastructure demands of large organizations.

“

Team8, the global venture fund, has announced $365 million in new capital, comprising a $265 million third fund and over $100 million earmarked for follow-on investments in its most promising portfolio companies.

The core Fund III will target Seed and Series A companies in cybersecurity, software infrastructure, fintech, and digital health. These sectors, according to Team8, are on the front lines of the AI revolution. The fund’s first exit occurred before its final close: Palo Alto Networks’ reported acquisition of portfolio company Koi Security in February 2026, signaling early validation of the investment thesis. Managing Partners Sarit Firon and Liran Grinberg, along with Partners Ori Barzilay and Hadar Siterman Norris, are leading the charge, bringing deep operational and investment expertise to bear.

The fundraising arrives at a critical inflection point. AI models from pioneers like OpenAI and Anthropic are evolving at breakneck speed, forcing enterprises to constantly rethink product architectures, workflows, and risk postures. Simultaneously, the rush to deploy agentic AI—autonomous software agents that can plan, reason, and execute tasks—has outstripped the implementation of adequate security controls. Team8’s own research at its CISO Village Summit, surveying 111 security leaders, revealed that 97% of organizations have already begun adopting AI agents, with 80% already running them in production. Yet the same survey found that confidence in their ability to secure these agents averaged a mere 2.32 on a scale of 1 to 5. This disconnect represents both a dangerous exposure for enterprises and a massive market opportunity for the right founders.

Team8’s capital will fuel startups that embed security and observability into AI-native tools from the ground up, rather than bolting on protections later. This aligns with the growing demand from CIOs and CISOs for solutions that provide visibility into agentic AI behavior, enforce governance policies, and safeguard data across complex software supply chains. The firm’s focus on enterprise sectors suggests a pragmatic thesis: AI will become the operating system of business, but its adoption will be gated by trust. Startups that can prove they are “AI-native” and “enterprise-ready” from day one will command premium valuations and rapid adoption.